The intuitive appeal of learning from the past to alter the present

Examining the past to better understand our present and future and even to change the way we act in the present and future, is something that has instinctive appeal. The argument is often made that by better understanding the context within which past harm took place, and the societal factors that contributed to it, we will better be able to recognize these circumstances when they re-emerge and be able to prevent harm from happening again in the future...

agencies and acts of violence in the everyday" may constitute "a continuity between the space of the prison and its outside". The ubiquitous presence of security-service kiosks or the routine assaults by security services on ordinary citizens going about their daily chores, for example, has blurred the boundaries demarcating the camp and the territory beyond it. 2 Ismail argues that in Syria today these camps, like the Sednaya and Adra prison, are indeed used in a manner that is aimed at breaking resistance, undoing the political subject, dehumanization, and even as part of a politics of annihilation. In such a context, Lyndsey Stonebridge refers to internment as marking the end of rights and, as such, a potential "prelude to genocide". 3 In this article I explore the way in which internment has been dealt with by several transitional justice mechanisms. This is important because these mechanisms have a "definitional" potential. As Zinaida Miller argues, "Transitional justice as both literature and practice offers more than just a set of neutral instruments for the achievement of the goals of justice, truth and reconciliation. It also serves to narrate conflict and peace, voice and silence, tolerable structural violence and intolerable physical atrocity. Ultimately, transitional justice is a definitional project, explaining who has been silenced by delineating who may now speak, describing past violence by deciding what and who will be punished." 4 I first zoom in on two cornerstones of the transitional justice architecture: trials and truth commissions, before turning to the broader domain of memorialization. Then I take a step back and reflect on what this means for how we think about justice and injustice.

The forward-looking role of backward-looking mechanisms
Transitional justice, as both a phenomenon and a conceptual tool, is regarded as inevitable and commonplace for anyone wishing to address the issue of past human rights violations. 5 In the domain of transitional justice an assumption of linear progress from a violent past towards a 2 Ibid., 194. more peaceful future is foundational, the logic being that we need to deal with that violent past -in a variety of ways -to arrive at a more just and peaceful future. 6 To this end, standardized interpretations of transitional justice encompass a range of practices and mechanisms, among other things, tribunals, truth commissions, reparations, institutional reform, constitutional reform, memorialization practices, and official apologies. Here I focus on tribunals and truth commissions, not only because of their importance in transitional justice, but also because they raise interesting questions regarding our positioning on a violent past. They do so because they contain both a backward-looking logic (for example, finding out about past harm in order to hold perpetrators accountable) and a forwardlooking logic which is aimed at preventing violence from happening again. 7 The growing importance of this forward-looking role means that a range of practices that was initially conceived of as mostly backwardlooking have been recast and reinterpreted in ways that also imbue them with a forward-looking role. This growing attention to forward-looking and preventive functions can be attributed to the fact that forty per cent of all countries where some or all of the measures in the transitional justice toolkit were implemented saw a relapse to violence within five years of signing peace agreements, making prevention a crucial consideration. Given the resources and institutional constraints of transitional justice mechanisms, it would be essential to focus on empowering citizen and former victims to take the justice process forward after the international actors leave, rather than "doing justice" in a retrospective manner within the liminal space of the transitional justice intervention.
The forward-looking function is most obvious in the case of trials, where a guilty verdict is argued as justified not only on retributive grounds ("it's the right thing to do", "because the perpetrator deserves it"), but also for its deterrent effect on future perpetrators. Retribution theories are backward-looking, deterrence theories forward-looking and expressivist: the verdict sends a message, as Mark Drumbl has written. 8 (Semi-)judicial proceedings can, however, also express messages in a more indirect manner. Carsten Stahn, in this context, calls international criminal justice a hallmark of social performance that creates social meaning and opportunities for shaping new narratives about past harm, and for reaching new audiences with messages about (what can be expected of) justice. Both deterrence and expressivist theories stand in opposition to views like that of Hannah Arendt that "the purpose of a trial is to render justice and nothing else". 9 This article inscribes itself in that expressivist turn in that it is concerned with the pedagogical value and legitimating qualities of justice processes. 10 Judicial proceedings play an expressive role when they apply and interpret laws and regulations. As such, they serve more than a legal purpose in that they also convey a moral judgment. Their extra-legal function then consists of sending an aspirational message about the norms and values that are deemed most relevant in a society, and also about what we consider to be outside the realm of justice. 11 This attention to expressivist functions of (semi-)judicial processes has also shaped the domain of transitional justice, which has increasingly been considered to be a definitional project. 12 Below I explore the extent to which these trials and truth commissions organized as part of a transitional justice process have been inclined to consider the issue of internment as asui generis -rights violation.

Trials
Following the end of a violent conflict, criminal justice proceedings are typically organized to hold perpetrators to account for past crimes. While having an obvious punitive function of establishing legal guilt, they also have the extra-legal expressive functions just described: often these processes play a role in later memorialization efforts, reparation orders, or even social mobilization or media reporting about the case. They play an important role in shaping our understanding of what is considered a crime and what is not. While legal practitioners of course have no monopoly over setting this historical record, it could easily be argued that, due to the status of these processes, they authoritatively shape the historical record, which may later influence other initiatives. Moreover, Lorraine Bannai, writing on Japanese-American internment, for example, has put forward the idea that an important motivation to go to court for many victims is that "The stories of those interned are worth remembering so that those who might be unfairly cast as the enemy today will not stand alone." 13 Randle DeFalco argues that these processes, and international criminal justice more generally, tend to prioritize crimes that are at the same time "spectacular and recognizable", and to pay less attention to more structural, indirect or socio-economic kinds of harm. 14 Internment, in that sense, is an issue which easily fits international criminal proceedings' parameters of legibility.
Yet, explicit references to internment as such are rare, with most transitional justice literature as well as court documents treating internment in a more general sense under the established legal categories of arbitrary or unlawful detention, not as a crime sui generis.
In the rulings of the Cambodian Khmer Rouge Tribunal, for example, there is exactly one mention of internment, when they refer to the intentional and arbitrary imprisonment without legal basis in the infamous Tuol Sleng S-21 prison, which was de facto an internment camp for -perceived -enemies of the party. In using the term "internment", the Tribunal acknowledged internment as a specific type of crime, different from related crimes such as other forms of arbitrary detention. Yet, at the same time, the court did not go so far as to seize the opportunity explicitly to condemn internment per se or in principle, instead ruling merely that there were no legal grounds or security requirements to justify detention in specific cases. The court structured its reasoning around the fact that babies and small children were among the detainees, who could objectively speaking not have posed a security threat or otherwise "have been guilty of any offence". Given how many people were tortured and detained in Tuol Sleng because they were believed to possess information or to be "enemies of the state", the court left an opportunity untapped by not investigating, or condemning, the situation and detention of those who were perceived as enemies of the state, or potential bearers of information.
Where court cases do explicitly revolve around internment, they often follow a logic of sequencing. This means that those crimes which are considered the gravest are dealt with first, and other issues are prosecuted at a later stage, often decades later. This has been the case in Chile, as well as, for example, in the United States. There, two cases were brought to court at the time that internment orders against Japanese-American citizens were being issued. These orders set in motion the mass transportation and relocation of more than 120,000 Japanese people to camps that were set up and occupied in about fourteen weeks. A Supreme Court ruling on Hirabayashi was issued in 1943, upholding the constitutionality of military curfew against a minority group as a military necessity. Another case that was brought to court in 1944 concerned the actual internment order. 15 But it was not until 1983 that there was a first judicial acknowledgment of the gravity of these rights violations, when the American Civil Liberties Union (ACLU) supported one of the plaintiffs, Fred Korematsu, and was granted the petition. 16 Also in this case there was no sweeping principled condemnation of internment as such. Instead, the court granted the petition because it acknowledged that the government had knowingly and willingly withheld full information on the case in 1944.
This case is also relevant because it is exemplary of the wished-for crossfertilization between backward-looking and forward-looking functions of legal proceedings: on the one hand, Fred Korematsu initially brought the case to court to seek legal accountability from those who had violated his rights (thus, a backward-looking goal and one related to his obvious personal stake in the matter). On the other hand, his legal mobilization of 1943-44 eventually led to his further legal and societal mobilization through his work in the National Coalition for Redress and Reparations. As Lorraine Bannai remarks, "At that point, Fred was transformed from someone who simply wanted to live free to someone committed to fighting the internment in court." 17 He wanted to see legislation to prevent these 15 See Peter Irons, Justice at War: The Story of the Japanese-American Internment Cases (Berkeley and Los Angeles: University of California Press, 1983). 16 Bannai, "Taking the Stand", 32. 17 Ibid., 10-11. kinds of issues from happening again. This is one case in which initial legal activism led to future human rights activism.
It is one of the few cases though where internment was explicitly dealt with and approached sui generis. Despite the magnitude of this injustice and its massive impact on the individual, the society, and the polity, a sequencing logic often pushes this crime down the line, and when criminal justice procedures do address it, it is often treated under the same established legal categories as other kinds of unlawful or arbitrary detention. This risks invisibilizing the fundamentally different nature and logic of internment, which is particularly worrying if we see these tribunals and courts as definitional instances with forward-looking and expressive functions. The question then arises of what kind of acknowledgment is taking place in the vast majority of cases that fail to treat internment sui generis. Later in this article, I return to this question by linking it to the notion of visibility politics and justice imagination.

Truth-seeking
First, however, I focus on another transitional justice mechanism whose primary function was also a backward-looking one, but which has increasingly been framed and lauded because of its alleged forwardlooking and preventive potential, namely truth commissions.
Nunca mas ("never again") reports emerged all over Latin America in the mid-1990s and early 2000s, following bloody dictatorships that often led to large-scale killings, disappearances, torture, displacement, and in some cases genocide. These reports had as their explicit goal to document and bring to the fore the nature and extent of human rights violations that took place to prevent them from ever happening again. It is interesting that in these reports the future is imagined in terms of the absence of something bad -Nunca más, nie wieder, never again -rather than in terms of the presence of something desirable, such as better protection of economic, social, and cultural rights, or the conceptualization of positive freedom. 18 Truth commissions are often seen as having a distinct role in writing what Michael Humphrey has called a "first draft of history", 19 in which victim memory is transformed into public knowledge through testimony. Such a view, Simon Robins argues, represents a highly positivist under standing of history, and indeed of the capacity for memory to be unambiguously corralled into a narrative that can effectively and uncontroversially serve a pre-determined purpose. 20 Similar to criminal justice proceedings, the mandates and reports of truth commissions tend not to mention internment explicitly, instead capturing it under more general sections regarding arbitrary arrests, inhumane treatment, or other broader references to unlawful detention. This again risks invisibilizing internment as a specific crime that is in its finality, perhaps not legally but analytically, different from the broader other types of crimes that it is framed as here.
The US is again an example of where a truth-seeking body was explicitly mandated to focus on internment. The Commission on War-Time Relocation and Internment of Citizens, created by Congress in 1982 to conduct a government study on the policies and effect of placing Japanese Americans in internment camps during the Second World War, was not a classic truth commission but a historical clarification commission. It was established with the intention of informing the public about these facts and providing reparations. As many as 750 people testified and there were several public hearings, which resulted in a report concluding that the internment of Japanese Americans had not been justified by military necessity and constituted a grave injustice. Most of its recommendations were implemented, including the passage of legislation, such as the Civil Liberties Act of 1988. This is also one of the few instances where a government apologized publicly and explicitly for the specific crime of internment, and where a public education fund and reparations revolved around, and named, the crime of internment specifically.
The Commission on War-Time Relocation and Internment of Citizens also shows that its forward-looking function lay in its concrete proposals, such as new legislation that would prevent the crime from happening again, and in its rhetoric that foregrounded internment and sought to raise awareness about its specificities as opposed to other detention crimes. None of the other truth commissions examined for this article so much as mention the term explicitly in their mandates, and hardly refer to it in their reports. This is striking considering that, as Zinaida Miller argues, "The mandate of a commission or the list of crimes to be tried at a court means that, to some degree, the decision about what story to tell is predetermined . . . The power to define the discourse of violation, the vocabulary of outrage and the appropriateness of dissatisfaction lies largely within the realm of transitional justice mechanisms, which deploy the human rights norms that originally constituted them. This power has been underestimated in many ways, and the repetition of particular crimes or certain indictable offences through both courts and commissions may have unexpected effects beyond the standard debates." 21 Miller's statement sheds light on how problematic it is not to acknowledge internment as sui generis in the mandates and reports of truth commissions. She also denounces the extent to which these commissions overlook the structural and socioeconomic violence and marginalization that often precedes more blatant rights violations by rendering whole groups of people vulnerable to gross violations of their rights. By omitting certain issues from the discourse, these bodies are, at least implicitly, defining what should be considered a crime and what will cause moral outrage.

A forward-looking memory
Narratives of past harm are, of course, not only shaped by (semi-)judicial bodies, which are not the only definitional interface between the past and the present. Other transitional justice mechanisms (as well as nontransitional justice processes, such as journalists' reports) also play a role in shaping these narratives. Here I concentrate on memorialization practices and memorial museums in particular, as these have been receiving growing attention in the domain of transitional justice because of their alleged forward-looking potential.
Memorial museums, for example, are increasingly promoted under transitional justice's banner of guarantees of non-recurrence, as their status moved from a mostly backward-looking function of documenting, archiving, and making information about past crimes publicly available, to a more forward-looking one of playing a role in challenging or even preventing human rights abuses that are taking place today. I discuss them under a separate heading here because their forward-looking function lies exclusively in their expressive potential to shape narratives (unlike courts whose forward-looking function can also lie in their deterrent function or option to issue reparation orders), and because establishing a historical narrative that can positively affect the present and the future is, in fact, a key concern for most memorial museums.
The growing importance of memorial museums, Huyssen argues, coincides with a broader societal interest in memory, memory sites, and memorialization. He calls this "the emergence of memory as a key concern in Western societies". Since the 1980s, many scholars have described this "boom" in both memory and memory studies. 22 Alongside the concern with memory writ large came a concern with and interest in remembering, memorialization, and remembrance education.
This can partly be understood by reference to our "civic duty to know" about the harms that have been inflicted in our names, the violations that have been carried out by our governments under the banner of guaranteeing our liberal democratic values, and the policies that have harmed others in order to guarantee our freedom. As Amy Sodaro puts it, "Memorial museums focused on past violence, atrocity, and human rights abuses reflect a demand today that those darkest days in human history are not only preserved but musealized and interpreted in a way that is widely accessible to present and future audiences." 23 From the start, however, this turn to memory was not only about the civic duty to know. Whether we consider historical memory education, memorials, or memorial museums, there seems to be an underlying assumption that well-designed and well-implemented interventions in the domain of remembrance provide opportunities to engage critically in comparative and prospective reflection, that, in other words, these remembrance and memorialization initiatives can provide lessons for the present and the future. In all these examples, the assumption is strong that history can, and indeed must, be used to shape not just our understanding but also our affects and actions; that either through careful historical analysis and comparison (for example, on the psychology of obedience, the dynamics of othering, the relation between structural and direct violence), or, in contrast, through a more emotive approach, we can plausibly impact the present and the future. There has, thus, always been an assumption that remembrance and memorialization can inform our behaviour in the present. That assumption also encompasses the aspiration of using history to challenge or even prevent today's human rights abuses.
Memorial museums, for instance, including the Japanese-American internment museum in Arkansas or the Tuol Sleng prison museum and memorial site just outside Phnom Penh, are meant to be institutions that morally educate their visitors, and promote human rights and an ethic of "never again". That this particular cultural form of commemoration is increasingly used globally as one of the central mechanisms for addressing past violence, Sodaro argues, suggests that it is believed to be an especially effective mode for critical engagement with the past that can translate into a more democratic and peaceful present and future. These memorial museums reflect the prevalent assumption that there is a causal relationship between learning about past violence and preventing it in the future. Not only do memorial museums intellectually educate their audiences about history, but they also seek to reach their visitors emotionally in order to transform them morally so that they embrace the ethic of "never again". Behind each museum is the claim that it is an essential part of building democratic culture and preventing future violence and atrocity through its creation of a more informed moral public that will work towards these goals. 24 According to the Colombian National Museum of Memory, for example, one of the core functions of memorial museums is to contribute to the construction of a culture of respect for difference, diversity, and plurality, or, as its mission statement puts it, "harnessing the memory and history of the violent past in a way that shapes the present and future". Such a description, however, hardly sheds light on the extent to which memorial museums, much like tribunals and truth commissions, are deeply political institutions, which are often created and utilized with specific political agendas that can (and often do) compromise their declared efforts openly to confront and learn from the past. Sodaro goes on to demonstrate, moreover, that memorial museums, rather than educating about the past, tend to reveal the political priorities and goals of the regimes that build them, reminding us that memory still resides in the domain of the nationstate, with the past being simply another arena for enacting present politics. 25 This concern over politicization becomes all the more relevant as remembering and memorializing violent pasts increasingly came to be seen 24 Ibid. 25 Ibid. as pathways towards more peaceful futures, and even as a precondition for democratic and inclusive societies. For one, (political) pressure increased for the state to be involved in these specific kinds of memory practices, the idea being that memorialization and remembering is an obligation. 26 More importantly, as remembrance and memorialization increasingly came to be seen as the sole focus of our engagement with tragedies of the past, the risk increased of the past becoming locked into itself, and memorialization becoming a largely stale conversation or symbolic ritual, rather than a lived understanding or practice. This challenges the possibility of drawing lessons for the future. 27 The kind of remembering that happens in such a context, moreover, does not always find it easy to rely on a version of history that is multi-vocal, multi-directional, multilayered, or otherwise complex or disruptive of our existing frameworks for thinking about the past, and more broadly for thinking normatively about what we condemn. Huyssen argues that what is at stake is the distinction between usable pasts and disposable data. 28 Yet when the transitory, incomplete, and complex nature of memorialization is brushed over, the version of the past that emerges is a curated "recognizable" one that does not easily accommodate cognitive or affective friction or dissonance, but instead serves as a basis for learning straightforward moral and political lessons about the present. Such a version of the past is often characterized by micro-studies that confirm our knowledge and views of the past, and that give us a feeling of "moral certainty and security". 29 This risks blurring the thin line that exists between "the real and the mythical past". 30 As Simon Robins posits, "acknowledging the importance of memory as a subjective record of the past alongside history, liberates history from sovereign power and drives an emancipatory approach in which history can in principle become something collaboratively created at multiple social levels". 31 We should, therefore, not only commemorate and memorialize the past, but also analyse how the past is invoked, so that we can better challenge its instrumentalization. This also raises questions about how lessons could be learnt, or how insights about the past might alter how we understand and act on our present-day realities. The intuitive appeal and normative imperative of memorialization and remembrance have often obscured the need to think systematically and analytically about how to ensure that these memory practices have the best chance of making any kind of impact on our understanding of, and actions in, the present and future. This question is particularly pertinent to historical memory education but, essentially, underlies all memory practices that seek to have a forwardlooking function. By historical memory education I mean programmes linking peace education, human rights education, democratic citizenship education, and Holocaust education to ask how we can take insights from past massive human rights violations in order to prevent them in the future.
One logic often cited in this regard is that critically engaging with the past nourishes other forms of critical thinking. But this requires precisely the kind of intellectual friction and disruption that is not easily accommodated in official (and often politicized) memorialization and remembrance programmes. Several state-of-the-art memorial museums, for example, are conceptualized in ways that seem to facilitate commemorative forms that can express a more ambiguous relationship to the violence of the past, but at the same time also clearly attempt to pass down pre-determined lessons, mostly about democratic values like freedom, tolerance, human rights, and the prevention of future violence. Yet, this second goal intuitively seems to clash with the open-endedness and ambiguity implied in the first.
Another logic revolves around the notion of empathy, and is pinned to the idea that memorialization, remembrance, and historical memory education can also instil a sense of empathy with victims of ongoing rights violations.
While both avenues are plausible, the theories of change or pathways of impact underlying them are markedly under-specified and we have little sound empirical evidence telling us which approaches work best under which circumstances. Should historical memory education focus more on cognitive aspects, on soliciting empathy with victims of abuses, or on empowering people to speak up or claim their rights? What do we see as a success in terms of, for example, historical memory education? And can we ever (even hope to) measure the effect of our efforts in this extremely complex area? If our concern with memorialization and remembrance revolves around the ambition of learning from the past to alter the future, we need urgently to answer these questions, and produce sound empirical research on how effects are likely to come about in a sustainable way. As Anja Mihr posited in 2015, if we want historical memory education to teach more than feeling empathy with the victims, and if it is to be more than simply a history lesson, then better measurement of impacts and more research on pathways of impact are needed. 32 The other option is to be more cautious about an overly instrumental or positivist interpretation of why remembrance and memorialization are important, and to revalue "our civic duty to know" and to engage with the past in critical ways that allow for cognitive and affective friction, and that do justice to multi-vocality, multi-directionality, and complexity. To address this question, I explore how narratives shape our justice imagination. 33

The narratives that shape our justice imagination
Earlier I argued that formal transitional justice interventions with an instrumental view of ensuring accountability for past harm, as well as remembrance and memorialization initiatives in the cultural domain, can be understood as definitional projects that narrate the violence of the past, and shape what causes moral outrage. I have also argued that attention to the specific dynamics of internment is virtually absent, especially from the more formal transitional justice mechanisms, which tend not to distinguish between the dynamics and intentionality of internment and those of other forms of unlawful detention. This is related to the difficulty for judicial and semi-judicial proceedings to account for experiences which are not formally codified in law. They create narratives from which the subject of internment is virtually absent or not acknowledged as sui generis. Such narratives narrow our understanding of the past, and at the same time influence what crimes will in future cause moral outrage.
In this last section, therefore, I focus on the issue of narratives and what they do to our understanding of harm and injustice. This is not to argue that narratives are the most important thing transitional justice interventions create. Indeed, as Siphiwe Dube has argued, these narratives, impactful as they are, "must always be accompanied by other forms of responses, such as structural changes and political commitments to social justice". 34 However, we should also acknowledge these narratives themselves as normative signifiers. Exploring how they operate in society can balance an overly positivist and instrumental reading of the past, as well as exploring how the past and present can interact in producing scenarios for the future without falling back into grand narratives. 35 This is not only a question of what we consider as harm and injustice, but also of how we narrate violence and harm, both within and beyond the legal arena.
I have already hinted at the importance of narratives from the standpoint of expressivist theories of the law and of justice processes. Narratives are also crucial if we consider that in each of the mechanisms described the process of having one's suffering acknowledged as a crime or as a gross violation of rights, as well as the struggle to find historical, legal, and political recognition, is essentially a testimonial one. Lyndsey Stonebridge has argued that if today we have so much memory just to start to work with, it is because witness testimony was placed at the heart of the judicial processes, starting with the Eichmann trial in 1961 and the trials following it. 36 She goes on to argue that Eichmann's trial was in fact the first time in history that a legal process was tied to claims of collective memory. And this starts to hint at the intersection between the various mechanisms of dealing with the past that I have been talking about: the witness testimonies provided during trials and the testimonies given during truth commissions (together with expert studies) often form the basis of further "musealization", and remembrance and memorialization practices, for example. Today most agree that it is through hearing the testimony of victims that the need for justice is felt most keenly. However, this practice seems to be grounded in an empathic imperative, an imperative to witness and absorb the suffering of the other; that raises the question of how soon before it falters, before the fantasy of the deserving and undeserving victim sets in if we too readily subscribe to narratives that brush over complexities. This question is particularly relevant when talking about the internment of people who not long before were cast as enemies of the nation.
Yet, this narrative style that presents itself as univocal and stable is what is preferred in both formal transitional justice mechanisms such as courts or truth commissions, as well as in more informal (or formal) memorialization and remembrance practices. This goes to the question of whether judicial rhetoric can ever aspire to have the capacity to allow for a testimonial style that is adequate to the subject who suffered unspeakable harm. Hannah Arendt, for example, argued that the real history of the camps must be told, but questioned whether this testimony could have a political-juridical value in a courtroom whose rhetoric and logic is notoriously ill-equipped to accommodate these stories and even these realities. This should not be read as a matter of locking suffering out of the political or judicial realm, nor as an attempt to make victims voiceless, but rather as an attempt to expose the limits of courtroom rhetoric, the politics of the use of the traumatized voice, and the perceived moral void at the heart of legal reasoning, which tends to be grounded in moral absolutism. 37 The difficulty and even impossibility of putting the experience of the camps into the language demanded by a court is a lamentation heard from Second World War victims to witnesses at the Khmer Rouge Tribunal or the International Criminal Tribunal for the former Yugoslavia (ICTY). Indeed, as Marie-Bénédicte Dembour and Emily Haslam show in their work on the ICTY, the more the classic language of the trial seeks to represent trauma, the more it is actually unable to connect with that reality -a reality of injury so grave that it cannot be expressed in the legal language at hand, but only articulated as traumatic experience, in a different moral realm altogether. 38 These arguments raise concerns about the limit of what can be understood about the injuries of others, especially within a courtroom. If the arguments are linked to the previous section about the expressive function of the law, they also raise concerns about what we perceive as harm and as injustice, and thus about what we imagine a more just future would look like.
Lyndsey Stonebridge's notion of "judicial imagination" is useful to start addressing this concern. She uses the notion to explore how we can imagine justice in cases of all-out trauma and violence, and how we can avoid falling into the legal phantasy that large-scale violence can be tried or addressed exclusively through the law. Building on this notion, I invoke the concept of "justice imagination", to also capture the role played by non-judicial mechanisms in shaping our faculty to generate new ideas 37 Ibid., discussing Arendt, 9. about how to curb injustices. And it might stretch the boundaries of what is conceivable in terms of justice, beyond present mechanisms and even beyond the judicial realm. 39 Such broader understanding would be a means to overcome the limits of legal reasoning. In a maximalist sense, a more ambitious justice imagination could refer to foraging for more encompassing justice narratives (either within or beyond existing justice mechanisms) that could accommodate the lived experiences of victims in a comprehensive manner. In a minimalist sense, it could refer to resisting the erasure or invisibilization of certain experiences from justice narratives that are heavily shaped either by legal language (which therefore do not treat internment as a specific category) or by stakeholders imposing their own (political) agendas onto memory practices, thereby rendering them less fruitful.
A crucial question to that end is whether and how we can move from trying to mould witness testimony about "the unspeakable", and also victims' aspirations for justice, into existing legal formats, towards a situation in which bearing witness can start to open up and push the boundaries of these legal avenues, to carve out space for a new language, a more ambitious justice imagination. This is not a plea for emotive narratives that solicit empathy, but instead for us to examine how we can allow for narratives that lack a stable referent or progressive linearity. Indeed, a language that better captures suffering, and thus solicits first an empathic reaction is likely not to facilitate the kind of analytical response that would be needed to understand this experience -or its causes in broader socio-economic and ideological conditions. What narrative mode is capacious enough, then, to open up the justice imagination on the basis of acknowledging lived experiences?
Here again Arendt can be an inspiration. In her discussion of Rahel Varnhagen's work, she proposes that what is needed is a poetic language, in the sense of a new language in which all the words have lost their banality, that is, their usual historical overuse. She refers to Varnhagen's idea that "only in the deliverance and total purity of poetry, which speaks every word as if for the first time, can she take language into herself, entrust herself to language . . . ungoverned by any tradition, any conventional norm." 40 Stonebridge summarizes Arendt's proposal as that judgments should not be just a matter of applying legal reasoning, 39 Herremans and Destrooper, "Stirring the Justice Imagination". 40 Arendt, cited in Stonebridge, Placeless People, 62. but should be reflective and imaginative, bringing our relationship with objects and concepts out into the open. 41 Along the same lines, Dube argues that "Fictional narratives are useful precisely because they offer polysemic descriptions of transitional moments, thereby bearing witness to these moments' complicated nature". 42 Both hint at the importance of reimagining the relation between language and justice, between narrative and judgment.
Where do we go from here?
The discussion here underlines the importance of interrogating our own role as a community of scholars and practitioners in creating and furthering narratives that leave entire dimensions of past human rights abuses undiscussed, and that can lead to invisibilization or erasure. This is important in the first place because of our "civic duty to know", but even more so because of the increasing emphasis on the forwardlooking function of transitional justice mechanisms in general, and memorialization and remembrance in particular. I have argued that much work remains to be done by those making a positivist claim about the instrumental value of memory practices in establishing the most plausible pathways of impact.
Yet, whether one starts from the civic duty to know or from a more instrumental concern with forward-looking effects, it is crucial to scrutinize what we explicitly define as an injustice and how we discuss those injustices. This is particularly important if one considers the expressive function of the law. Contrary to Arendt's argument that trials should not be treated as awareness-raising tools or as instances that write an authoritative historical account, but merely as establishing legal guilt, I have emphasized here the vast extent to which these judicial and semijudicial mechanisms in practice have an expressive function and do serve as a definitional project, bracketing off what comes to be seen later as an injustice. This brings us back to the issue of internment -and to the question of learning from the past to prevent violations from taking place in the future.
Japanese Americans who were interned have received token compensation and a formal apology for their years of internment. Lorraine Bannai argues, however, that they have not received what many of them really 41 Ibid. 42 Dube, "Transitional Justice Beyond the Normative", 195. sought: assurance that a similar injustice will never occur again. This was their demand, and was explicitly inscribed in Title 6 of the Civil Liberties Act of 1988. 43 In fact, in 2019 some commentators referred to internment once again, as news from the US-Mexican border shocked the world. The US was reported to have about 30,000 people in what is euphemistically called "administrative immigration detention", casting these individuals as criminals who pose a threat to the security of the nation, justifying their detention as a preventive measure, and excluding them from access to civil or human rights. Seventy-five years after the internment of Japanese Americans, the country was vilifying, and the government was detaining, groups of individuals based on their ethnic affiliations. Suspicion was cast on individuals who "looked like" the enemy. The reference to internment triggered a fierce debate about whether these camps could be equated with internment camps. This debate sometimes got stranded in definitional battles that obscured the extent to which, as Stonebridge argues, historically there is nothing unusual about these kinds of camps becoming other kinds of camps -internment camps, for example. 44 Both present-day and past examples of this kind of violation of human rights should not be invisibilized, neither narratively nor in the practice of and mobilization for justice. To grasp fully their complexity, we need to adopt a language -in legal forums and beyond -that allows for the acknowledgment of their complexity and transient nature, as well as accounting for their embeddedness in dynamics of othering, dehumanization, and societal polarization, and structural and cultural violence, which often precedes more gross crimes and human rights violations.